Roman Population Theories

This study is getting a ton of press today, which I'm finding kind of interesting due to the relatively obscure nature of the topic and findings. I'd love to hear what the UNRV community thinks of this.

Here are links to a number of different stories.

from Physorg.com: Buried Coins Key to Roman Population Mystery?

from Live Science: Ancient Rome's Real Population Revealed

from Wired: Roman Coin Hoards Show More War Means Fewer Babies

The new work could help settle a long-standing historical debate about the Roman population. Census figures from the end of the second century B.C. show a population of adult males of around 400,000. Then, the record goes blank, and census figures from around a hundred years later show a population of 4 to 5 million. Some of the population explosion is explained by the extension of Roman citizenship to various groups, but far from all of it. From this evidence, a group of historians known as the

This study is getting a ton of press today, which I'm finding kind of interesting due to the relatively obscure nature of the topic and findings. I'd love to hear what the UNRV community thinks of this.

Among the many excellent posts on archaeological findings by our JG, I think this has been the best for some time; it may even become a watershed in the demographic studies of Ancient Rome, because it provides quite hard evidence for the "Low Count" hypothesis, which should presumably be already considered the standard explanation, as it is in fact implicitly assumed by most reviewers.

Some significant consequences of the "Low Count" are that:

- the count of the Roman Censuses shifted from the adult males to the whole population under Augustus, even if unattested by our available sources (a good example of a positivist fallacy);

- the Roman urbanization was atypically high for the first century BC;

- their military draft was tremendous for any place or time.

This study is getting a ton of press today, which I'm finding kind of interesting due to the relatively obscure nature of the topic and findings. I'd love to hear what the UNRV community thinks of this.

I suppose that the relatively "obscure nature" is in the eye of the beholder; I have been a big fan of the work of Prof. Scheidel for some time.

Among the many excellent posts on archaeological findings by our JG, I think this has been the best for some time; it may even become a watershed in the demographic studies of Ancient Rome, because it provides quite hard evidence for the "Low Count" hypothesis, which should presumably be already considered the standard explanation, as it is in fact implicitly assumed by most reviewers.

 

Some significant consequences of the "Low Count" are that:

- the count of the Roman Censuses shifted from the adult males to the whole population under Augustus, even if unattested by our available sources (a good example of a positivist fallacy);

- the Roman urbanization was atypically high for the first century BC;

- their military draft was tremendous for any place or time.

Sylla,

Yeah...I guess I was editorializing a little bit, eh? Clearly from the amount of coverage this study is receiving, there's interest in the topic.

Question for you...what would be the implication if the "High Count" theory had resulted? Each article referenced "rewriting Roman history", but none that I saw provided much rationale.

Question for you...what would be the implication if the "High Count" theory had resulted? Each article referenced "rewriting Roman history", but none that I saw provided much rationale.

"According to the researchers, mapping out the times when the coins were buried is a good indirect method for measuring the intensity of internal warfare and unrest, /// and therefore a key indicator of population demographics"

I agree with the first part but I can not see how it's related with the second. Having more hoards means a more intense and widespread conflict rather then a larger population.

"According to the researchers, mapping out the times when the coins were buried is a good indirect method for measuring the intensity of internal warfare and unrest, /// and therefore a key indicator of population demographics"

I agree with the first part but I can not see how it's related with the second. Having more hoards means a more intense and widespread conflict rather then a larger population.

On the face of it this seems to be an important study however, I would tend to agree with Kosmo that while the incidences of hoards are a good indicator of areas under some form of stress they may simply indicate areas subject to population movement rather than necessarily the 'death' of the indiivisual(s) who buried the hoard(s).

Statistics is often seen as an arcane art but modelling population decline on this basis is in my view pushing the bounds of what can or cannot be proven to any degree of certainty. As I understand it one of the major concerns in any statisitical analysis is that you are using 'real' differences as your baseline, which given the lack of accurate records from this period leaves this study open to question. The authors seem to have started from the premise that there would be a congruance of population size to the number of hoards rather than is there any other way of modelling ancient populations - that second question seems to remain unanswered.

"According to the researchers, mapping out the times when the coins were buried is a good indirect method for measuring the intensity of internal warfare and unrest, /// and therefore a key indicator of population demographics"

I agree with the first part but I can not see how it's related with the second. Having more hoards means a more intense and widespread conflict rather then a larger population.

 

On the face of it this seems to be an important study however, I would tend to agree with Kosmo that while the incidences of hoards are a good indicator of areas under some form of stress they may simply indicate areas subject to population movement rather than necessarily the 'death' of the indiivisual(s) who buried the hoard(s).

 

Statistics is often seen as an arcane art but modelling population decline on this basis is in my view pushing the bounds of what can or cannot be proven to any degree of certainty. As I understand it one of the major concerns in any statisitical analysis is that you are using 'real' differences as your baseline, which given the lack of accurate records from this period leaves this study open to question. The authors seem to have started from the premise that there would be a congruance of population size to the number of hoards rather than is there any other way of modelling ancient populations - that second question seems to remain unanswered.

To defend this research I'd argue that internal turmoil and warfare rarely comes without a decrease in population. Furthermore, the researchers do not talk of hoards being linked to death but rather to population demographics (there being less people in Rome); wether they died or moved away is not the issue. The researchers simply linked the distribution of coin hoards to population demographic, that is population growth or decrease, not to the direct cause of it. When calculating population growth, one has to take into account four major factors: death, birth, immigration and emmigration. What the researchers are saying is that the amount of coin hoards proves widespread internal turmoil and warfare (they have proven this by applying their model to other times and places) and that widespread internal turmoil and warfare leads to a decrease in population. They do not specify but logically, during war and turmoil, more people die, less babies are born, less people move into the region and more people move out of the region; leading to population decrease or at the very least a serious slowdown of its growth.

"According to the researchers, mapping out the times when the coins were buried is a good indirect method for measuring the intensity of internal warfare and unrest, /// and therefore a key indicator of population demographics"

I agree with the first part but I can not see how it's related with the second. Having more hoards means a more intense and widespread conflict rather then a larger population.

 

On the face of it this seems to be an important study however, I would tend to agree with Kosmo that while the incidences of hoards are a good indicator of areas under some form of stress they may simply indicate areas subject to population movement rather than necessarily the 'death' of the indiivisual(s) who buried the hoard(s).

 

Statistics is often seen as an arcane art but modelling population decline on this basis is in my view pushing the bounds of what can or cannot be proven to any degree of certainty. As I understand it one of the major concerns in any statisitical analysis is that you are using 'real' differences as your baseline, which given the lack of accurate records from this period leaves this study open to question. The authors seem to have started from the premise that there would be a congruance of population size to the number of hoards rather than is there any other way of modelling ancient populations - that second question seems to remain unanswered.

 

To defend this research I'd argue that internal turmoil and warfare rarely comes without a decrease in population. Furthermore, the researchers do not talk of hoards being linked to death but rather to population demographics (there being less people in Rome); wether they died or moved away is not the issue. The researchers simply linked the distribution of coin hoards to population demographic, that is population growth or decrease, not to the direct cause of it. When calculating population growth, one has to take into account four major factors: death, birth, immigration and emmigration. What the researchers are saying is that the amount of coin hoards proves widespread internal turmoil and warfare (they have proven this by applying their model to other times and places) and that widespread internal turmoil and warfare leads to a decrease in population. They do not specify but logically, during war and turmoil, more people die, less babies are born, less people move into the region and more people move out of the region; leading to population decrease or at the very least a serious slowdown of its growth.

Latium antiquum a Tiberi Cerceios servatum est m. p. L longitudine: tam tenues primordio imperi fuere radices. colonis saepe mutatis tenuere alii aliis temporibus, Aborigenes, Pelasgi, Arcades, Siculi, Aurunci, Rutuli et ultra Cerceios Volsci, Osci, Ausones, unde nomen Lati processit ad Lirim amnem. in principio est Ostia colonia ab Romano rege deducta, oppidum Laurentum, lucus Iovis Indigetis, amnis Numicius, Ardea a Dana

As I cannot see the full research my comments are necessarily constrained to the press articles which on my first reading seemed to indicate they were comparing the first century BC male population of 2-400,000 with the 1st century AD total population of 4-5,000,000 of Roman and stating this was a 200 percent difference which it patently isn't.

On second reading it appears that I picked on the wrong article to reference as the other articles give more of an indication that they were talking about the difference in the first century BC estimates and using the incidence of coin hoards only as an indication of political instability rather than a 'causal link' which could be precisely quantified to population per se. I apologise if my misreading accidentally reversed what this study was intended to show.

BTW Sylla I tend to be more cynical when it comes to there only being one possible answer to every question which is asked or theretical model which is perfect - there are often varying degrees of probability. Technically it is not only 'dead people [who] were effectively prevented from recovering' hoards there are any number of other factors which could have prevented recovery including to name just a few; being involved in the long period of internal strife and having left Rome with one or other of the armies. Being disabled during the fighting could also prevent return to recover hidden property as would being sent into exile, being captured and held abroad/ being enslaved or put into prison or even either having something built over the site or nearby construction having a landmark removed making recovery uncertain to say the least.

Latium antiquum a Tiberi Cerceios servatum est m. p. L longitudine: tam tenues primordio imperi fuere radices. colonis saepe mutatis tenuere alii aliis temporibus, Aborigenes, Pelasgi, Arcades, Siculi, Aurunci, Rutuli et ultra Cerceios Volsci, Osci, Ausones, unde nomen Lati processit ad Lirim amnem. in principio est Ostia colonia ab Romano rege deducta, oppidum Laurentum, lucus Iovis Indigetis, amnis Numicius, Ardea a Dana

Besides of not being supported by the available evidence, your reflections are frankly naive, to say the least.

Now you are being naive - it is well known in archaeologiocal and numismatic circles that when it comes to hoards you need to look at all of the evidence of the internal composition of the hoard and not simply in isolation as this study appears to have done.

There are several distinct versions of hoard composition the main types being:

Savings hoards which tend to be long term deposits which are gradually added to by individuals or families.

'Regular' or emergency hoards which reflect the copinage in use of a particular period

Ritual hoards - normally deposited in a grave or at a temple site.

The internal chronology of the coin sequences, amount of wear, how many of a particular and/or recent issue are included in the hoard, even the number of 'fakes' are all important factors numismatically in determining which type of hoard is being considered and possible reasons for it not being recovered.

Changes to the purity of coinage as well as occupation by different regimes can be factors but death is NOT the only reason for non-recovery of hoards which have been recorded as regular deposits in other hoards and later recovery would have continued throughout any period of instability. The question of why a hoard was not recovered is obviously important and war can be a major factor while many of the issues outlined above are to a great extent subjective rather than precisely quantifiable so often do not appear in a quantifiable way in hoard reports. There obviously will have been variations in both the overall number of hoards in operation at any given period and the number in percentage terms actually recovered.

For all of these reasons I remain to be convinced that ALL of the hoards involved in this study have been fullly categorised in the way outlined above so they could be appropriately factored into the proposed model.

One way of assessing the validity of the coin hoards as a proxy for political violence is to compare it to other proxies like prosopographical evidence. PP and I did a little study a while back looking at all the named victims of political violence in the literary sources, and I made a chart to illustrate the findings.

Here's our chart (133 BCE - 44 BCE):

Now here's the chart from PNAS (200BCE - 0 CE):

The two sources--numismatic and prosopographic--align rather nicely. Note especially the two spikes surrounding the Marius/Sulla conflict (92-80) and Caesar's civil war (49-44). Moreover, the Italian hoards provide evidence of political turmoil (surround the Social War) that the prosopographical evidence neglects.

In my view, the PNAS article was superb and is a model of what modern ancient history ought to be.

One way of assessing the validity of the coin hoards as a proxy for political violence is to compare it to other proxies like prosopographical evidence. PP and I did a little study a while back looking at all the named victims of political violence in the literary sources, and I made a chart to illustrate the findings.

 

<SNIP>

In my view, the PNAS article was superb and is a model of what modern ancient history ought to be.

There does appear to be a degree of correlation but I am uncertain if it really answers my question about the reliability of the 'perceived' underlying 'date of closure' (when the hoard was last accessed and coins added) of the hoards used in the research as that is the factor, which from my reading of the articles, does not appear to have been fully addressed by this study. The articles are unclear but I suspect the study may for convenience have accepted the last datable coin as the date of closure of the hoard, however the physical condition of such coins (amount of wear) should also be considered as this can have implications for the 'actual' as opposed to 'perceived' date of closure of the hoard.

e.g if the last datable coin was actually a badly worn coin of 53BC this may well indicate that it was in circulation for many years after 53BC. Conversely a group of closely dated 53BC 'new' or at least only partly worn coins would make a 53BC date of closure for the hoard much more certain.

It comes down to a matter of correct interpretation of the underlying dating evidence also being an essential aspect of such research and this caveat should be noted when considering, what I agree, are the potentially positive aspects of this model.

Latium antiquum a Tiberi Cerceios servatum est m. p. L longitudine: tam tenues primordio imperi fuere radices. colonis saepe mutatis tenuere alii aliis temporibus, Aborigenes, Pelasgi, Arcades, Siculi, Aurunci, Rutuli et ultra Cerceios Volsci, Osci, Ausones, unde nomen Lati processit ad Lirim amnem. in principio est Ostia colonia ab Romano rege deducta, oppidum Laurentum, lucus Iovis Indigetis, amnis Numicius, Ardea a Dana

Latium antiquum a Tiberi Cerceios servatum est m. p. L longitudine: tam tenues primordio imperi fuere radices. colonis saepe mutatis tenuere alii aliis temporibus, Aborigenes, Pelasgi, Arcades, Siculi, Aurunci, Rutuli et ultra Cerceios Volsci, Osci, Ausones, unde nomen Lati processit ad Lirim amnem. in principio est Ostia colonia ab Romano rege deducta, oppidum Laurentum, lucus Iovis Indigetis, amnis Numicius, Ardea a Dana